On Sun, 14 Dec 2008, Marc Chantreux wrote:

On Sat, Dec 13, 2008 at 01:34:44PM +1100, Timothy S. Nelson wrote:
On Sat, 13 Dec 2008, howard chen wrote:
        What I think is more likely to happen in reality is that people will
make various Perl6 "distros", ie. the Perl6 core + whatever modules are
suitable to the purpose.  So we could have, for example:

.... and confuse the users by adding more names in the perl6 world. There
is allready too much words about perl6. Please concider that a newbie
will hear words like PGE,perl6,rakudo,pugs,parrot,PIR,PASM ...

Hopefully no newbie will hear of PIR and PASM once the final Perl6 is out. And I assume that the distros are easier than worrying about 1001 modules.

The next point, for me, is that distributions and bundles are just
useless: there are a lot of modules that comes with de standard
distribution i never used and i imagine that no sysop can imagine what
are *my* needs as netop:

- do i need to generate pdf reports to my boss?
- does my servers deal with apple or windows world?
- what kind of applications are running on my servers?

Well, when I reinstall my machine every few years, there are a few software packages that I download and install. But that's easier than going with Linux from Scratch :). I'm not suggesting that this will be a magic bullet any more than Fedora, Debian, or Ubuntu is a magic bullet.

depending to the answers, a lot of modules that would be very important
for one sysop will be useless for me.

when i compare my expects to my observations of others computers users,
i thinks that the best way to distribute perl6 and make it popular is to
provide a minimal installation with a powerfull tool to discover and
install perl6 modules (as dpkg/aptitude/synaptic does).

searching by tags, description, names, using cpan as we use google ...

Why bother creating a separate tool? Let's have something that automatically feeds CPAN into repositories for apt, yum, and whatever.

        Btw, in case you haven't seen it:


        If you haven't seen that, then you'll want to bookmark this:


Diamonds: Perl6 for the web distro (Web people make mony, buy diamonds)
Clubs: Perl6 for systems administrators (club users, before they club you)
Spades: Perl6 for gravediggers (??? :) )

        Naturally, the "Diamonds" mentioned above would include a particular
templating system, some XML support, modules for various useful
protocols, and the like.

- my sysop tasks requires template systems, do i have to install diamonds?
 (i think that you think about install Clubs and adding templating
 modules but this is confusing for new people)

Well, if I were designing Clubs, I'd have put templating in there already, but if that wasn't what had happened, then you'd do exactly as you said; install Clubs, plus a templating system.

- my minimal set for Diamonds contains LMTP,IPP,XMPP,... but a lot of
 web developpers don't deal with them .. so what to choose ?

Hmm. Well, while I think most web developers would want something that does e-mail (LMTP), the others would be, as you said, things that people would add on to Diamonds.

        I guess I can see the value in many of the existing templating systems,
but would like to see a single one designed in a "postmodern" fashion
that can accomodate the supporters of all existing templating systems.

this could be called Utopia!

        Quite possibly.  See my other post for more details.

| Name: Tim Nelson                 | Because the Creator is,        |
| E-mail: wayl...@wayland.id.au    | I am                           |

Version 3.12
GCS d+++ s+: a- C++$ U+++$ P+++$ L+++ E- W+ N+ w--- V- PE(+) Y+>++ PGP->+++ R(+) !tv b++ DI++++ D G+ e++>++++ h! y-

Reply via email to